And here's the back of them. The baffle is made out of 2.5" acrylic, reinforced with 1/4" aluminum angles and pressure fit mounted between wooden studs (on 4 spike plates). The gap between baffle and wooden frame is filled with expandable foam.
This is a wall between 2 rooms. To make it more silent, the hole inside cavity of this wall is filed with sand, and 3/4 MDF is used on both sides, glued to studs with industrial adhesive and additionally screws were used.
There is only one capacitor in tweeter network, no crossover on Triangle and one coil limiting PHL driver from the top. No binding posts, speaker cable soldered directly to driver posts.
This is a wall between 2 rooms. To make it more silent, the hole inside cavity of this wall is filed with sand, and 3/4 MDF is used on both sides, glued to studs with industrial adhesive and additionally screws were used.
There is only one capacitor in tweeter network, no crossover on Triangle and one coil limiting PHL driver from the top. No binding posts, speaker cable soldered directly to driver posts.
Attachments
Re: LT versus LM
Though the LT part has better specs in your nice graphs I wouln't be surprised if the LM part is the better sounding.......
Hi Rick,Rick NL said:
Though the LT part has better specs in your nice graphs I wouln't be surprised if the LM part is the better sounding.......
Re: Re: LT versus LM
Why?
Elso Kwak said:
Hi Rick,
Though the LT part has better specs in your nice graphs I wouln't be surprised if the LM part is the better sounding.......
Why?
Peter, I apologize, your parting line read as a smart remark, I see now it wasn’t.
I don’t use the buffer or an inverting topology, It would be interesting to read your views of the regulated supply without the buffer, this may be an easier comparison as there are probably too many variables with the extra circuitry that goes with adding a buffer.
Neal
I don’t use the buffer or an inverting topology, It would be interesting to read your views of the regulated supply without the buffer, this may be an easier comparison as there are probably too many variables with the extra circuitry that goes with adding a buffer.
Neal
I have been out this afternoon but just had time before my evening meal to power up the buffered, regulated IGC.
First comment: very low DC offset at around 18 mV for both channels. This is the lowest that I have achieved with a straight wire from NI to ground!
I have a buzz coming from my test speaker but it goes when the buffer section is powered down. I have done a quick and sloppy hook up of the attenuator so I will look at that side next.
BUT - playing some music even through my test speaker indicates there is an improvement over the un-regulated buffered IGC! Very clear and smoooooth with clearly defined bass. Basically Gainclone sound but more so if you know what I mean.
More football tonight so I will continue work tomorrow! 😉
First comment: very low DC offset at around 18 mV for both channels. This is the lowest that I have achieved with a straight wire from NI to ground!
I have a buzz coming from my test speaker but it goes when the buffer section is powered down. I have done a quick and sloppy hook up of the attenuator so I will look at that side next.
BUT - playing some music even through my test speaker indicates there is an improvement over the un-regulated buffered IGC! Very clear and smoooooth with clearly defined bass. Basically Gainclone sound but more so if you know what I mean.
More football tonight so I will continue work tomorrow! 😉
Nuuk said:
I believe it is indicative of some rail decoupling problems. a small electrolytic and a mica / ceramic capacitor should do.
Rick,
Linear gives two columns of the data, one describing LT and the second describing the LM part, but all the graphs in this datasheet applies both to the LT and LM part. So my conclusion is there is no difference between these two units in the parameters described by the graphs and the differences you pointed out should be understood like differences between National’s LM338 and Linear’s both LM338 and LT338. Some of these differences may be the consequences of possibly different measurement setups/conditions, what is known is National tested its unit under 2A load, but this is just a guess. The only clear thing is that the worse looking output impedance of the National’s part above a few tens of kHz is caused by the fact the National setup did not include an output cap of 10uF as Linear’s.
Pedja
Linear gives two columns of the data, one describing LT and the second describing the LM part, but all the graphs in this datasheet applies both to the LT and LM part. So my conclusion is there is no difference between these two units in the parameters described by the graphs and the differences you pointed out should be understood like differences between National’s LM338 and Linear’s both LM338 and LT338. Some of these differences may be the consequences of possibly different measurement setups/conditions, what is known is National tested its unit under 2A load, but this is just a guess. The only clear thing is that the worse looking output impedance of the National’s part above a few tens of kHz is caused by the fact the National setup did not include an output cap of 10uF as Linear’s.
Pedja
Nuuk said:
Don't tell me this is better because I've just thrown my amp out of the window!!!
Gotta go get the pieces!!!
See ya later!
😎
Pedja said:
Hi Pedja,
Yes Linear has both in their program they state:
The LT138A is an improved version of the popular LM138 with improved circuit design and advanced process techniques to provide superior performance and reliability.
The text in their Datasheets further only refers to the LT.
It is not clear by this that both Linerar LT and Linear LM are the same in relation to the graphs. Also it is certainly not sure that the National LM and the Linear LM are the same.
I only have experience with the OP37 and the LT1037 where the LT is clearly superior on sound.
I or you or someone else have to compare the three to be sure. Then you could compare the three devices on their own or in the regulated gainclone, where you have what you finally want but don't be sure what changed parameter is contributing to a change in sound.
The world does not get less complex, when you know more.
Rick
regulated PS
I am using 100 VA 2x25 regulated this is more than enough
with my relatively inefficient speakers and a large room 20 x 28 and high ceiling. I have some heavy Bach pipe organ it sounds marvellous.
This is my second GC I am retesting with LM338 Regulators. I have no voltage sag even with small Xfrmr. This was not the case with 400VA and large caps at PS PCB
The sound is definitely better than my first GC with Plitron unregulated with 400VA at 2x25, which is destined to become 4 channel amp.
My conclusion is that at least with regs, 100 VA is more than enough for 2 channels.
I have: 2x 2000 uf at the PS PCB and 2x 5600 uf right after. I have reduced cap to 47 uf at the 3875, bypassed with 0.1 ufd poly. the regulator has 10 uf Electro on the adjust pin and a 1uf tantalum right on the reg input with Carlos.47 on the output. Both diodes on the reg i have them on heat sink and T03 package which is likely overkill. Using 120R and 3k0 for resistor s on the regulator.
Even with small Xfrmr no voltage sag during heavy bass passages. I am dropping 4 volts from 36.7 to 32.6 volts.
So reducing the chip capacitor size and lots of caps after rectifiers seems the way to go.
So my vote goes to regulated, I have not buffered (yet?)
I am not convinced yet of the need. I only have CD input.
But the improved power supply is a definite plus and I will be working on the other amp again.
SheldonD
I am using 100 VA 2x25 regulated this is more than enough
with my relatively inefficient speakers and a large room 20 x 28 and high ceiling. I have some heavy Bach pipe organ it sounds marvellous.
This is my second GC I am retesting with LM338 Regulators. I have no voltage sag even with small Xfrmr. This was not the case with 400VA and large caps at PS PCB
The sound is definitely better than my first GC with Plitron unregulated with 400VA at 2x25, which is destined to become 4 channel amp.
My conclusion is that at least with regs, 100 VA is more than enough for 2 channels.
I have: 2x 2000 uf at the PS PCB and 2x 5600 uf right after. I have reduced cap to 47 uf at the 3875, bypassed with 0.1 ufd poly. the regulator has 10 uf Electro on the adjust pin and a 1uf tantalum right on the reg input with Carlos.47 on the output. Both diodes on the reg i have them on heat sink and T03 package which is likely overkill. Using 120R and 3k0 for resistor s on the regulator.
Even with small Xfrmr no voltage sag during heavy bass passages. I am dropping 4 volts from 36.7 to 32.6 volts.
So reducing the chip capacitor size and lots of caps after rectifiers seems the way to go.
So my vote goes to regulated, I have not buffered (yet?)
I am not convinced yet of the need. I only have CD input.
But the improved power supply is a definite plus and I will be working on the other amp again.
SheldonD
That's interesting Sheldon! Particularly about not needing a large VA transformer.
As regards the buffer I think it is needed for a few reasons but only if there is no other active circuit ( ie preamp) between source and IGC.
First, it is necessary IF you want to try an LPF so that the LPF is isolated from the previous stages/cables etc.
Second, in my experience the GC's sound better when they are driven by an active preamp or buffer.
Third, the buffer isolates the input of the IGC preventing any problems with long cables affecting the input loop.
I have quite a bit of experience of buffered IGC's but am considering going for a preamp (including active crossover etc) and regulated unbuffered NIGC's.
As ever, with comparatively few parts involved, the individual should try all these variations for themselves and choose the combination that sounds best in their system. 😉
As regards the buffer I think it is needed for a few reasons but only if there is no other active circuit ( ie preamp) between source and IGC.
First, it is necessary IF you want to try an LPF so that the LPF is isolated from the previous stages/cables etc.
Second, in my experience the GC's sound better when they are driven by an active preamp or buffer.
Third, the buffer isolates the input of the IGC preventing any problems with long cables affecting the input loop.
I have quite a bit of experience of buffered IGC's but am considering going for a preamp (including active crossover etc) and regulated unbuffered NIGC's.
As ever, with comparatively few parts involved, the individual should try all these variations for themselves and choose the combination that sounds best in their system. 😉
Off topic but as a Greek national I can not resist.
Greece 1 - Czech Republic 0 The fantastic spirit of the greek players did another miracle and beat one of the favorites....Our journey continues....
Carlosfm - our teams meet again and hopefully we will be able to repeat our first match result.....
Cheers,
Kosta
Greece 1 - Czech Republic 0 The fantastic spirit of the greek players did another miracle and beat one of the favorites....Our journey continues....
Carlosfm - our teams meet again and hopefully we will be able to repeat our first match result.....
Cheers,
Kosta
reg PS
further to previous testing:
the smaller Xfrmr now sags at the AC level but there is no sag at the 3875 PCB. This is because the LM338 is doing a fine job of regulating the output voltage to the chips.
This is in stark contrast to 440 VA Plitron on with large caps inPS but unregulated. There was no sag on the AC input to powersupply but even with 2x 5600uf per rail but the pana 1500uf on the chip pcb, there was sag of 3% on lloud passages with heavy bass.
So it looks like regs could make up for undersized Xfrmrs.
So as a second conclusion multi channel GC could savea lot by
having 100VA as a base, then multiply by number pof channels. I,E, having one 600VA say for 6 channels, especially with music having differen needs on different channels at the same instant.
SheldonD
further to previous testing:
the smaller Xfrmr now sags at the AC level but there is no sag at the 3875 PCB. This is because the LM338 is doing a fine job of regulating the output voltage to the chips.
This is in stark contrast to 440 VA Plitron on with large caps inPS but unregulated. There was no sag on the AC input to powersupply but even with 2x 5600uf per rail but the pana 1500uf on the chip pcb, there was sag of 3% on lloud passages with heavy bass.
So it looks like regs could make up for undersized Xfrmrs.
So as a second conclusion multi channel GC could savea lot by
having 100VA as a base, then multiply by number pof channels. I,E, having one 600VA say for 6 channels, especially with music having differen needs on different channels at the same instant.
SheldonD
Hi Rick,Rick NL said:
I am aware of it, but it is only that "improved circuit design and advanced process techniques to provide superior performance and reliability" tells me nothing. While at the other hand I could distinguish the designer's goal was apparently to achieve more predictable ouput voltage and I concluded it is about voltage reference. Another fact that made me think so is that there are no other visible signs of changed parameters.
Pedja
Nuuk said:
I as joking, of course.
It's that only one member says this is no good, and I know what's his problem.😉
Nuuk said:
Nuuk, I've said that, pay attention to the posts...right on the first pages of this thread, and look at the pics.
This trafo I scavenged from an old amp is much smaller than the big toroid I had in my unregulated GC.
But this thing really kicks.
It's obvious that a regulated PSU with big caps before the regs is much better than an unregulated PSU, even with a big toroid, with 1000~1500uf caps.
If you go higher you muck up the sound, so no other option.
Anyone has any doubt that this reg thing has MUCH BETTER BASS?
Say it again?
Oh... only one.
I knew it...
exipnos said:
This time we are prepared.😀
I don't know if you noticed, but we changed half the team since the first game with Greece.😉
Pay attention? Well, with so many arguments in this thread stretching it to 80+ pages, it's a bit difficult to remember just what was said on those first pages! 😉
I have just been reading the first first three pages again but can't find a reference to the VA of your transformer.
But no matter. The point is I was just replying to something Sheldon said in his post! Let's not make a Krell out of a Gainclone. 😉
- Status
- Not open for further replies.